Method and device for analyzing and evaluating radar data of radar sensor

Abstract

In a method for analyzing and evaluating radar data of a radar sensor, the radar sensor generates individual radar data in a plurality of measurement periods. The radar data is subdivided and assigned to a range of parameters of the radar spectrum. Radar data associated with the range of parameters of the radar spectrum is stored for a measurement period dependent on the number of the range of parameters. For each parameter range, the radar data associated with the parameter range is evaluated. The invention also relates to a device for analyzing and evaluating radar data of a radar sensor.

Description

TECHNICAL FIELD

[0001] The present invention relates to a method and a device for an evaluation of radar data of a radar sensor. BACKGROUND

[0002] Different methods are known for modulating a radar signal. The modulated radar signal is emitted, reflected on an object, and the reflected radar signal is received again in order to determine the distance, the relative speed and / or the angular position of the object.

[0003] A known modulation method is the so-called Chirp-Sequence method, in which a fast frequency ramp or a chirp packet with a duration T f is emitted. Subsequently, a pause T P is provided. The chirp packet with the subsequent pause corresponds to a measurement cycle with a duration T f + T P .

[0004] An alternative method for radar modulation is the Orthogonal Frequency Division Multiplex (OFDM), in which the bandwidth BW is sampled by means of a plurality of orthogonal subcarriers. The time sampling is carried out by emitting a plurality of so-called OFDM symbols.

[0005] From US 2020 / 0408879 A1 a device for an evaluation of radar data of a radar sensor is known. SUMMARY

[0006] The present invention provides a method and a device for an evaluation of radar data of a radar sensor according to the invention.

[0007] Preferred embodiments are subject matter of the respective dependent claims.

[0008] According to a first aspect, the invention relates to a method for an evaluation of radar data of a radar sensor. The radar sensor generates individual radar data in a plurality of measurement cycles. The radar data is subdivided and attributed to a parameter range of a radar spectrum. The radar data attributed to the parameter range of the radar spectrum is stored for a number of measurement cycles depending on the parameter range. For each parameter range, the radar data attributed to the parameter range is evaluated.

[0009] According to a second aspect, the present invention relates to an apparatus for analyzing and evaluating radar data from a radar sensor, the apparatus having an interface for receiving radar data generated by the radar sensor in multiple measurement cycles. A computing facility subdivides the radar data and assigns these subdivided radar data to parameter ranges of the radar spectrum. The radar data assigned to the parameter ranges of the radar spectrum is stored in a memory for measurement cycles depending on the number of these parameter ranges. The computing facility analyzes and evaluates the radar data assigned to each parameter range.

[0010] Advantages of the present invention

[0011] In radar measurements, the range resolution (Δd) depends on the bandwidth (BW) used. The relative velocity resolution (Δv) depends on the measurement duration (T). f :

[0012]

[0013] Here, c represents the speed of light and f0 represents the center frequency of the radar modulation. To achieve maximum range, or effective range d, without undersampling... max Within this range, precise distance measurements must be performed, and during equidistant sampling, the distances must be distributed across the bandwidth (BW).

[0014] N = d max / Δd

[0015] Each sample value. To measure the relevant relative velocity range [v] min ,v max To perform a precise measurement of relative velocity, during equidistant sampling, the measurement duration T must be specified. f Upper distribution

[0016] M = (v max -v min ) / Δv

[0017] Each time sample value.

[0018] Maximum measurement duration T in a single package f It may be subject to limitations, for example, due to requirements regarding the maximum clearly measurable relative velocity, the maximum permissible delay until output radar positioning, or thermal requirements.

[0019] By jointly analyzing and evaluating radar data from multiple measurement cycles (e.g., multiple linear frequency modulated signal packets), the effective measurement duration can be significantly increased, thereby significantly improving the signal-to-noise ratio and relative velocity resolution compared to a single measurement cycle.

[0020] For a common evaluation of a plurality of measurement cycles, the radar data of all measurement cycles must be stored jointly in the memory of the radar sensor or must be transmitted when the data are processed on a central control device, which places certain demands on the transmission speed and the digital memory.

[0021] The method according to the application reduces the memory requirement for the evaluation across cycles. To this end, for each parameter range, the radar data are stored for a respective predefined number of measurement cycles. Here, the data assigned to a first parameter range can be stored, for example, for a longer time than the radar data assigned to a second parameter range. As a result, the radar data assigned to the second parameter range can be deleted again earlier and the memory requirement is thereby reduced.

[0022] The application therefore finally proposes a radar data-dependent, targeted storage of the parameter ranges of the radar spectrum.

[0023] The memory can be partitioned, for example, in such a way that performance advantages are retained for the relevant application scenarios and at the same time the amount of data used for the calculation for the evaluation across cycles is significantly reduced. The radar data (radar frames) can be stored, for example, only for regions of the radar spectrum (or radar image) in which the evaluation across cycles brings advantages for the relevant application scenarios. These radar data are then stored in the memory and evaluated across measurement cycles. In this way, the amount of memory required can be substantially reduced.

[0024] According to a further embodiment of the method for the evaluation of radar data of a radar sensor, the radar data for each parameter range of the radar spectrum are stored in a first-in first-out memory, i.e. a FIFO memory. Here, the memory depth depends on the parameter range.

[0025] According to a further embodiment of the method for the evaluation of radar data of a radar sensor, the parameters of the parameter range comprise at least one of the distance, the relative speed, the azimuth angle and the elevation angle. Each parameter is then assigned a respective value in the radar spectrum, for example the scattering cross section or the amplitude.

[0026] According to a further embodiment of the method for the evaluation of radar data of a radar sensor, the radar data assigned to a first distance range are stored for a greater number of measurement cycles than the radar data assigned to a second distance range. Here, the second distance range is closer to the radar sensor than the first distance range. A delay can occur when the evaluation across measurement cycles is carried out, but this delay has less of an effect for distance ranges for more distant distances, so that more measurement cycles can be taken into account there.

[0027] According to another embodiment of the method for the evaluation of radar data of a radar sensor, radar data assigned to a first angular range in the center in front of the radar sensor are stored for a greater number of measurement cycles than radar data assigned to a second angular range at the side. The central angular range is usually of particular importance.

[0028] According to another embodiment of the method for the evaluation of radar data of a radar sensor, partial information of the assigned radar data is selected and stored for at least one parameter range. For example, specific parameter ranges can be defined in which there are usually statistical or due to feedback from the environment recognition objectives which are not identifiable in the evaluation based on only real-time measurement cycles, but which can be identified in the evaluation across measurement cycles. It can thereby be ensured that as many objectives as possible are present in the stored radar data which can be identified by the evaluation across measurement cycles.

[0029] Here, the splitting of the parameter ranges can be performed according to the statistical occurrence probability of the objectives, for example, subdivided by distance and velocity or by angle. The splitting of the parameter ranges can also be performed according to fixed relationships between the detected objectives, for example, according to the number of detected objectives per distance range. The splitting of the parameter ranges can also be performed according to limitations due to transmission speed and / or available memory. Another possibility for the splitting of the parameter ranges is to take into account feedback to the environment recognition, for example, by scene recognition.

[0030] According to another embodiment of the method for the evaluation of radar data of a radar sensor, the selection of the partial information of the radar data is performed according to the ratio of the signal power to a constant false alarm rate threshold, i.e., a CFAR threshold.

[0031] According to another embodiment of the method for the evaluation of radar data of a radar sensor, the subdivision of the parameter ranges of the radar spectrum is dynamically generated. By the dynamic generation, further improvements in memory efficiency can be achieved.

[0032] According to another embodiment of the method for the evaluation of radar data of a radar sensor, the number of measurement cycles assigned to a parameter range is dynamically determined. Thereby, the memory efficiency can be improved.

[0033] According to another embodiment of the method for the evaluation of radar data of a radar sensor, the dynamic generation of the subdivision of the parameter ranges and / or the determination of the number of measurement cycles assigned to a parameter range is performed according to the movement of the radar sensor. In particular, the ego trajectory of the radar sensor (or of the motor vehicle in which the radar sensor is located) can be taken into account.

[0034] According to another embodiment of the method for the evaluation of radar data of a radar sensor, the dynamic generation of a subdivision of parameter ranges and / or the determination of a number of measurement cycles assigned to a parameter range is performed in dependence on the evaluation of the radar data. In particular, the evaluation of parameter ranges in which a relevant object has been identified can be performed over a plurality of measurement cycles, i.e. the number of measurement cycles can be increased.

[0035] According to another embodiment of the method for the evaluation of radar data of a radar sensor, the evaluation of the radar data comprises a combination of radar data detected in a measurement cycle. Thereby, for example, the resolution can be improved.

[0036] According to another embodiment of the method for the evaluation of radar data of a radar sensor, the radar sensor operates according to a chirp sequence method or an OFDM method. BRIEF DESCRIPTION OF DRAWINGS

[0037] Further advantages, features and details of the present application result from the subsequent description, in which different embodiments are explained in detail with reference to the drawings.

[0038] The accompanying drawings show:

[0039] Figure 1 Schematic block diagram of a radar sensor with an apparatus for the evaluation of radar data of a radar sensor according to an embodiment of the present application;

[0040] Figure 2 One exemplary subdivision of distance parameter ranges / velocity parameter ranges of a radar spectrum;

[0041] Figure 3 Another exemplary subdivision of distance parameter ranges / velocity parameter ranges of a radar spectrum;

[0042] Figure 4 Another exemplary subdivision of distance parameter ranges / velocity parameter ranges of a radar spectrum;

[0043] Figure 5 One exemplary subdivision of azimuth angle parameter ranges of a radar spectrum;

[0044] Figure 6 Schematic diagram for illustrating the storage of radar data;

[0045] Figure 7 Flow chart of a method for the evaluation of radar data of a radar sensor according to an embodiment of the present application.

[0046] In all figures, elements and devices that have the same or similar function are designated with the same reference signs. The numbering of the method steps is for clarity and should not imply a specific chronological order unless otherwise indicated. Several of the method steps can also be performed simultaneously. DETAILED DESCRIPTION

[0047] Figure 1 A schematic block diagram of a radar sensor 1 is shown, which has a device 5 for the analytical evaluation of radar data of the radar sensor 1. The device 5 can be part of the radar sensor 1 or external.

[0048] The device 5 comprises an interface 2, which receives radar data produced by a transmitter / receiver facility 6 of the radar sensor 1 in a plurality of measurement cycles. The radar data are saved in a memory 3. A computing facility 4 subdivides the radar data and attributes these subdivided radar data to parameter ranges of a radar spectrum.

[0049] For this purpose, the computing facility 4 can first perform a radar processing of the radar data by known methods for determining distance and velocity, for example by a formation of a two-dimensional Fast Fourier Transform (FFT). Optionally, an additional angle estimation can additionally be performed on the complete radar spectrum or in individual parameter ranges, for example by beamforming or using a Deterministic Maximum Likelihood (DML) estimator.

[0050] The computing facility 4 takes the radar spectrum. Parameters of the radar spectrum are, for example, distance, relative velocity, azimuth angle, elevation angle or a selection (subset) of the aforementioned parameters.

[0051] The radar data attributed to the parameter ranges of the radar spectrum are stored in the memory 3 for a number of measurement cycles depending on the parameter range. For example, the radar data can be stored in a first-in-first-out memory, i.e. a FIFO memory 3, for each parameter range of the radar spectrum.

[0052] The radar spectrum is thus divided into parameter ranges for which different memory depths are defined, since the number of measurement cycles for which the radar data are stored in a particular parameter range depends on the parameter range itself.

[0053] The computing facility 4 analytically evaluates the radar data attributed to each parameter range. For this purpose, the radar data stored over the respective number of measurement cycles can first be combined for each parameter range.

[0054] Here, a Multi-Frame Integration (MFI) method can be used, such as a Chirp-Sequence-3D method, a Keystone method or a Backprojection method. In this way, a plurality of high-resolution partial radar images can be obtained by combining radar data of a plurality of measurement cycles.

[0055] The parameter range of the radar spectrum can be statically subdivided, i.e. the parameter range of the radar spectrum can be fixedly predefined. For example, the static subdivision can be made in accordance with requirements of typical driving situations. For example, for targets at a large distance, it can be advantageous to analyze and evaluate over a large number of measurement cycles in order to improve the signal-to-noise ratio of otherwise very weak targets and in order to achieve a high speed resolution, so that otherwise possibly inseparable targets can be separated in terms of speed, for example.

[0056] For example, due to the ego speed of the radar sensor, a Doppler shift occurs in the case of static objects. By means of the ego speed v ego is projected onto a static target having a velocity

[0057] v Ziel = v ego cos θ Ziel

[0058] where θ Ziel denotes the azimuth angle of the target, in particular only a small Doppler shift occurs for targets close to the boresight, i.e. in the center in front of the radar sensor, which can be solved with a larger number of combined measurement cycles

[0059] In another embodiment, the radar data assigned to a first speed range, which corresponds to targets moving towards the radar sensor, can be stored over a larger number of measurement cycles than the radar data assigned to a second speed range. Such targets are usually more important than targets moving away from the ego radar sensor or vehicle, for example in the second speed range, so that an analysis and evaluation over a plurality of measurement cycles is advantageous for the first speed range, whereas an analysis and evaluation over one measurement cycle can be sufficient in the second speed range.

[0060] The subdivision of the parameter range of the radar spectrum can be made dynamically. Additionally or alternatively, the determination of the number of measurement cycles assigned to a parameter range can be made dynamically.

[0061] The dynamic generation of the subdivision of the parameter range and / or the determination of the number of measurement cycles assigned to the parameter range can be performed in accordance with a motion of the radar sensor. For example, the division of the parameter range can be based on a determined ego trajectory of the radar device. Depending on the velocity and the driving direction, different regions in the distance spectrum, the velocity spectrum and / or the angle spectrum are of interest. For example, at high ego velocities, targets at long distances are more important than at low ego velocities.

[0062] Additionally or alternatively, the dynamic generation of the subdivision of the parameter range and / or the determination of the number of measurement cycles assigned to the parameter range can be performed in accordance with an analysis evaluation of the radar data.

[0063] In a subsequent processing step of the perception, for example, objects can be identified and their trajectories estimated based on the radar targets and possibly further information sources, such as other sensors or map data. In this process, specific parameter ranges of the radar spectrum can be identified, in which radar measurements with a higher signal-to-noise ratio or a higher Doppler resolution can lead to a better perception of the situation.

[0064] For example, it can be identified that the ego vehicle is approaching a bridge, wherein the memory depth for the multi-cycle processing is increased in the respective parameter range of the radar spectrum so much that the velocity resolution is improved such that the bridge can be separated early at the end of a possible jam.

[0065] Figure 2 An exemplary division of the distance (d) parameter range / velocity (v) parameter range of the radar spectrum is shown, in which exemplary values for the memory depth are noted. Here, the radar spectrum is subdivided into different parameter ranges depending on the distance d and the velocity. Here, a greater memory depth is assigned to at least some parameter ranges with greater distance, i.e. the number of measurement cycles is, for example, 4 or 5 instead of 1 to 3: The radar data of the respective parameter range is stored over said measurement cycles. However, these values are to be understood only exemplarily. Thus, in this example, the radar data assigned to the first distance range can be stored for a greater number of measurement cycles than the radar data assigned to the second distance range. Here, the second distance range is closer to the radar sensor 1 than the first distance range.

[0066] Figure 3 Another exemplary division of the distance parameter range / velocity parameter range (d-v diagram) of the radar spectrum is shown. The subdivision of the distance parameter range / velocity parameter range is performed in accordance with the ratio of the signal power of the radar radiation of the respective parameter range to a constant false alarm rate threshold, i.e. a CFAR threshold.

[0067] In the first parameter range 401 the signal-to-noise ratio of a single analysis evaluation is too low to identify a specific target, e.g. a pedestrian, with sufficient probability. For example, there the signal power is smaller than the CFAR threshold. However, an analysis evaluation across measurement cycles can provide detection if necessary, so that a large amount of data should be obtained there. Therefore, in the first parameter range 401 (spectrum) the overall information can be transmitted and stored. This information contains for example the complex amplitudes of each relevant transmitter / receiver antenna combination at each point inside this first parameter range 401.

[0068] The uncertainty is smaller in the second parameter range 402, but a possible occlusion can occur. Here, the signal-to-noise ratio can be sufficient for a single analysis evaluation in general, e.g. the signal power is greater than or equal to the CFAR threshold, with a small difference, e.g. less than 3 decibel. However, due to the nature of the scene or the nature of the distribution of targets, an occlusion of other targets can occur, e.g. in the velocity direction. Therefore, the radar data is selected so that the amount of data can be significantly reduced.

[0069] The second parameter range 402 comprises a section close to the detected target. There, the radar data is transmitted and stored together over multiple measurement cycles. The radar data can be selected, e.g. the complex amplitudes of each relevant transmitter / receiver antenna combination can be stored. Thereby, additional targets that would otherwise be occluded can be identified in an analysis evaluation across measurement cycles by the increased Doppler resolution capability.

[0070] For selecting relevant radar data, a detection can be performed first, e.g. using CFAR and subsequent peak detection. Thereby, a fixed or dynamically defined area around the detection can be selected. For example, a pre-defined number of pixels, e.g. 2 pixels, around the detection point can be selected statically, respectively. The number of pixels can also be dynamic, e.g. depending on the signal-to-noise ratio.

[0071] In the third parameter range 403 the signal-to-noise ratio of a single analysis evaluation is sufficient, e.g. the signal power is significantly larger than the CFAR threshold, and the probability of an occlusion due to the scene or the distribution of targets is almost excluded. In this case, for example the phase information, e.g. the complex amplitudes, of the detected target and all relevant transmitter / receiver antenna combinations can be transmitted and stored together. Alternatively, meta information, e.g. the peak width, can be added in addition to the typical parameters of a target detection.

[0072] Figure 4Another exemplary partitioning of the range parameter range / velocity parameter range of the radar spectrum is shown. On the left side the minimum CFAR threshold S is plotted as a function of the distance d. In the middle an exemplary illustration of the radar spectrum with the distance d and the relative velocity v is shown. On the right side the radar spectrum after the CFAR analysis evaluation is illustrated, wherein the radar data is partially or completely discarded in the parameter range 50. The CFAR threshold can generally be adjusted depending on the distance d, the relative velocity and / or the angle or also dynamically by feedback on the environment recognition.

[0073] Figure 5 An exemplary partitioning of the azimuth angle parameter range of the radar spectrum is shown, wherein exemplary values for the memory depth are noted. Targets located directly in front of the vehicle and thus significantly more likely to be located on the driving path are in some cases more worthy of attention than targets in the edge region. By way of the relationship

[0074] v Ziel = v ego cos θ Ziel

[0075] A greater difference in the projected Doppler shift is furthermore generated for targets with a greater azimuth angle offset, so that a coarser Doppler resolution capability can already be sufficient. Thus, in the present embodiment, a higher memory depth (e.g. 5) is assigned to the region located directly in front of the vehicle (in the forward direction B; "boresight") than to the azimuth angle ranges located further out (e.g. memory depth 1 or 2).

[0076] Figure 6 An illustration for illustrating the storage and processing of partial radar images over a plurality of measurement cycles is shown. Here, the radar data is subdivided and assigned to the respective parameter ranges, i.e. partial radar data or partial radar images 21 to 23 are generated, wherein three partial radar images 21 to 23 are exemplary plotted.

[0077] The respective most up-to-date partial radar images of the respective parameter range are stored in the respective FIFO memory 31, 32, 33. Here, the oldest entry, i.e. the oldest partial radar image 21, 22, 23, is discarded. Here, the depth of the FIFO memory 31, 32, 33 depends on how many measurement cycles are used to calculate the respective partial radar image 21, 22, 23. The high-resolution partial radar images 41, 42, 43 are calculated after each measurement cycle, independently of this depth. Thus, the depth of the FIFO memory 31, 32, 33 only influences the delay with which the complete, high-resolution partial radar image 41, 42, 43 is calculated, but not the update rate. When generating the high-resolution partial radar image 41, 42, 43, all partial radar images 21, 22, 23 in the respective FIFO memory 31, 32, 33 are combined and jointly evaluated. Finally, the high-resolution partial radar image 41, 42, 43 flows into the final radar image 50.

[0078] Figure 7 A flow chart of a method for evaluating radar data of a radar sensor, for example the radar sensor 1 described above, is shown.

[0079] In a first step S1, the radar sensor 1 generates respective radar data in a plurality of measurement cycles.

[0080] In a step S2, the radar data are subdivided and attributed to parameter ranges of a radar spectrum.

[0081] In a step S3, the radar data attributed to the parameter ranges of the radar spectrum are stored for a number of measurement cycles depending on the parameter range. The radar data can be stored in a first-in-first-out memory, i.e. a FIFO memory 3, for each parameter range of the radar spectrum. Here, the parameters of the parameter ranges can comprise at least one of a distance, a relative velocity, an azimuth angle and an elevation angle.

[0082] Here, the parameter ranges of the radar spectrum can be subdivided statically or also dynamically. The number of measurement cycles attributed to the parameter ranges can also be determined statically or dynamically.

[0083] The subdivision of the parameter ranges and / or the determination of the number of measurement cycles attributed to the parameter ranges can be dynamically generated depending on a motion, for example a trajectory, of the radar sensor 1.

[0084] Additionally or alternatively, the subdivision of the parameter ranges and / or the determination of the number of measurement cycles attributed to the parameter ranges can be dynamically generated depending on an evaluation of the radar data.

[0085] For each parameter range, the radar data assigned to this parameter range are evaluated. For this purpose, the radar data stored over a number of measurement cycles depending on the parameter range can be combined first.

Claims

1. A method for the evaluation of radar data of a radar sensor (1), the method having the following steps: generating (S1) individual radar data by the radar sensor (1) in a plurality of measurement cycles; The radar data are subdivided and attributed (S2) to parameter ranges of the radar spectrum, wherein storing the radar data assigned to a parameter range of a radar spectrum for a number of measurement cycles depending on the parameter range; evaluating (S3) the radar data assigned to the parameter range for each parameter range.

2. The method of claim 1, wherein, The radar data for each parameter range of the radar spectrum are stored in a first-in, first-out memory, FIFO memory (3).

3. The method of claim 1 or 2, wherein, The parameters of the parameter range include at least one of a distance, a relative velocity, an azimuth angle and an elevation angle.

4. The method of any of the above claims, wherein, The radar data assigned to a first distance range are stored for a greater number of measurement cycles than the radar data assigned to a second distance range, wherein the second distance range is closer to the radar sensor (1) than the first distance range.

5. The method of any of the above claims, wherein, The radar data assigned to a first angle range in the center in front of the radar sensor (1) are stored for a greater number of measurement cycles than the radar data assigned to a second angle range on the side.

6. The method of any of the above claims, wherein, Selecting and storing partial information of the radar data assigned for at least one parameter range.

7. The method of claim 6, wherein, The selection of the partial information of the radar data is made depending on a ratio of a signal power to a constant false alarm rate threshold, CFAR threshold.

8. The method of any of the above claims, wherein, The subdivision of the parameter ranges of the radar spectrum and / or the number of measurement cycles assigned to the parameter ranges are generated dynamically.

9. The method of claim 8, wherein, The dynamic generation is made depending on a movement of the radar sensor (1).

10. The method of claim 8 or 9, wherein, The dynamic generation is made depending on the evaluation of the radar data (1).

11. The method of any of the above claims, wherein, The evaluation of the radar data includes a step of combining the radar data stored over a number of measurement cycles depending on the parameter range.

12. An apparatus (5) for the evaluation of radar data of a radar sensor (1), the apparatus having: an interface (2) configured to receive radar data generated by the radar sensor (1) in a plurality of measurement cycles; a computing facility (4) configured to subdivide the radar data and to assign it to parameter ranges of a radar spectrum; a memory (3) configured to store the radar data assigned to a parameter range of the radar spectrum for a number of measurement cycles depending on the parameter range; wherein the computing facility (4) is further configured to evaluate the radar data assigned to the parameter range for each parameter range.

Images